Digital audio content can be communicated over networks in various different formats. When content is available in multiple formats, each of the available formats is often communicated over the network. In a broadcast or multicast context, communicating each of the available formats may be preferred, so that various receiving devices can select an appropriate format based on their respective configurations or capabilities. In a unicast or individualized context, however, communicating multiple audio formats may unnecessarily consume network resources. Accordingly, a need exists for device-based audio-format selection.